"Made in..." usually relates to where they were packaged and tested rather than a silicon statement.
This chip could have won plentiful of design wins, laptop manufacturers would have been happy to use it, not to mention customers.
Sometimes I think that there is something going on in that company and shoot themselves in the foot and no one else is responsible for their money bleeding processes.
What do you mean by in sight? Kabini is modern GCN it just doesn't have the same HSA capabilities as Kaveri.It's not only that. AMD threw Hainan/Oland/Sun in market segments, where Cape Verde/Bonaire should have been. NV uses a far broader range of GPUs for the mobile market, by downclocking. For OpenCL developement there are very few good mobile devices, because NV hangs up with CUDA. In my opinion its just a shame, because it looks like AMD is not even trying to compete in the mobile sector. Others might say, go with an APU, but a modern GCN APU is not even in sight, right now.
What do you mean by in sight? Kabini is modern GCN it just doesn't have the same HSA capabilities as Kaveri.
It's not only that. AMD threw Hainan/Oland/Sun in market segments, where Cape Verde/Bonaire should have been.
Of course, but Nakai mentioned OpenCL development not graphics performance. You could do a lot of OpenCL development with Kabini, just not performance tweaking if your target market isn't a Kabini class chip.Kabini offers graphics performance far below any modern discrete mobile GPU products.
Kaveri could compete with the discrete low-end, but as Nakai said the release of mobile Kaveri chips has yet to be given a date.
Of course, but Nakai mentioned OpenCL development not graphics performance. You could do a lot of OpenCL development with Kabini, just not performance tweaking if your target market isn't a Kabini class chip.
Maybe. When a thread is read over multiple days it's easy to forget what was previously discussed. :smile:One sentence of his post mentioned OpenCL development, but I was of the assumption that his post as a whole was in regards to overall graphics performance.
http://pci-ids.ucw.cz/read/PC/10026900 Topaz XT [Radeon R7 M260]
6901 Topaz PRO [Radeon R5 M255]
6920 Tonga
http://videocardz.com/50472/amd-launch-new-flagship-radeon-graphics-card-summer
more interesting, in my opinion:
- 28nm TSMC 1H 2014 : Hawaii, VI 1.0
- 28nm GlobalFoundries 2H 2014 : Iceland and Tonga, VI 2.0
- 28nm GlobalFoundries 1H 2015 : Maui, VI 2.0
- 20nm GlobalFoundries 2H 2015 : Fiji and Treasure, PI 1.0
- 20nm GlobalFoundries 1H 2016 : Bermuda, PI 1.0
- 14nm GlobalFoundries 2H 2016 : Mid-GPU and Low-GPU, PI 2.0
- 14nm GlobalFoundries 1H 2017 : High-GPU, PI 2.0
Difference shouldn't be big, ACEs are just buffers if I understand currectly. They help a few percent in asynchronous compute tasks but there really isn't much difference in measurable performance between the 2 even in compute driven tasks. There would be even less influence in games.I have a question about the ACEs/Queues of Southern Islands vs. Volcanic Islands: How are 8 ACEs with 64 Queues (VI) vis-a-vis 2 ACEs/2 queues (SI) going to influence performance of future games? I'm talking about game engines that have been optimized for XBO/PS4 (no PS3/360 cross-platform) and then ported to PC. Will the difference be decisive enough to make an upgrade necessary despite using moderate settings? What exactly will the ACEs influence? Thanks.
Is there a list of all "Islands", or can you tell please what island groups correspond to what gen? I've never even worked out what "Southern Islands" and "Northern Islands" were, especially due to the cancelled 32nm generation which was one of those orginally.